Ion implant is a critical technology in the fabrication of semiconductor devices. Ion implanters are typically used for performing ion implantation processes. The ion implanters are used to provide doping for the semiconductor devices, wherein impurity atoms are introduced to change the electrical properties of semiconductor materials.
A typical ion implantation process includes the steps of creating ions of the desired impurity atoms, using electric fields to accelerate the ions to a required energy, transporting the ions down a beam line to a silicon wafer, and scanning the beam, or moving the silicon wafer, or both, such that uniform dosage in the silicon wafer is accomplished.
All semiconductor fabrication processes use many (typically 15 to 25) steps of ion implantation to create a completed semiconductor device. The primary parameters of ion implantation are species, energy and dose. The species are the types of atoms being implanted, and there are two main categories, N-type and P-type, which are denoted by the electrical activity of the impurity in the semiconductor materials. N-type dopants are usually arsenic or phosphorus and P-type dopants usually include boron. The energy determines how deep into the silicon wafer the ions will go: high energy implants are deep, while low energy implants are shallow. The dose determines how conductive the implanted region will be. All of these parameters are chosen by the transistor designer for each implant step to optimize the device characteristics.
A basic ion implantation system is schematically shown in FIG. 1, which includes an implanter comprising an ion beam generator, a beam manipulation unit, and a process chamber. In the ion beam generator, plasma of the desired implant species is produced, from which ions are extracted. The resulting ion beam may then be passed through a magnet (not shown), so that ions of a particular mass to charge (m/e) ratio are selected. The ion beam that emerges from the magnet is further accelerated during the beam manipulation stage. The ion beam is then projected on the silicon wafer in the process chamber.
Since ions are electrically charged, and only ions with a certain number of charges are selected in a given implantation process, by determining the charges landing on the wafer, the ion dosage can be determined. The dosage determination is typically performed by a dose integrator electrically connected to the wafer, wherein the dose integrator determines the accumulated dosage implanted on the wafer. A schematic diagram of a dose integration process performed by a dose integrator is shown in FIG. 2, wherein a wafer current, which is generated due to the charges carried by ions, is input into the dose integrator and integrated with respect to time. The accumulated charge amount is converted to an accumulated dosage and is used to control the implanter.
In a typical implantation process, a target dosage is predetermined. The dose integrator determines the accumulated dosage by multiplying the current flowing through the dose integrator by the duration for which the current flows. The accumulated dosage is then compared to the target dosage. As soon as the target dosage is reached, the implantation process stops.
The dose integrator typically comprises common circuit components such as resistors, capacitors and operational amplifiers. Therefore, with time, the circuit components may degrade and the detected accumulated dosage may drift from the actual value. As a result, the implanted dosage will deviate from the predetermined value. For example, assume one coulomb of charges is expected by the dose integrator to provide the target dosage, and the charges are supplied by one-ampere current flowing through the dose amplifier for one second. If the dose integrator drifts, and the one-coulomb charges are wrongfully determined to be 0.8 coulombs, the implantation will then last 1.25 seconds instead of one second to make up the difference. As a result, 1.25 coulomb charges actually pass the dose integrator, which means that the corresponding dosage is also 25 percent more than necessary. This will cause degradation of, or even failure of, the resulting integrated circuits on the wafer.
Although dosage integrator drift does not occur frequently, when it occurs, the respective cost is high. Conventionally, periodic monitoring is the only way to catch this problem. However, periodic monitoring cannot catch the problem in real time, and many wafers may be damaged during the period of time from when the drift occurs to when the problem is found by e.g., daily monitoring. For example, up to about 1000 wafers may be manufactured and damaged due to the drift of the dosage integrator. Therefore, a method for catching dosage drift in real time to prevent high loss is needed.